Fix README: move Ko-fi up, add YAML metadata

README.md

@@ -25,13 +25,21 @@ RYS-enhanced variant of the Ornstein-3.6-27B dense model. Layer 33 is duplicated
 
 We are a proudly Canadian research collective working to advance **sovereign Canadian AI** — open-weight models that Canadians (and everyone else) can run locally, study, and build on without dependence on closed foreign APIs. All training, fine-tuning, and quantization is done on local and self-funded compute. By supporting this work, you help keep frontier model development accessible, transparent, and under Canadian stewardship.
 
-## Running Locally
+## Important: requires a patched llama.cpp
 
-This model requires a **patched llama.cpp** to load correctly. RYS breaks the hardcoded `full_attention_interval = 4` assumption in stock llama.cpp.
+RYS duplicates one of the middle layers, which breaks the hardcoded `full_attention_interval = 4` assumption in stock llama.cpp's Qwen3.5 loader. This model is converted with **per-layer `head_count_kv` baked in**, and you need a llama.cpp that reads that per-layer metadata instead of falling back to the interval formula.
 
-**Use this patched fork:** https://github.com/DJLougen/llama.cpp/tree/rys-qwen35
+**Patched fork:** [https://github.com/DJLougen/llama.cpp](https://github.com/DJLougen/llama.cpp) (default branch `rys-qwen35`, fully backward-compatible).
 
-The fork now includes both per-layer `layer_types` support and an **SSM tensor probing fallback**, so even legacy GGUFs load correctly. It is fully backward-compatible with non-RYS Qwen3.5 models.
+Stock llama.cpp, Ollama, LM Studio, and any other inference runtime built on stock llama.cpp will currently fail to load this model with a `check_tensor_dims` error — this is expected until/unless the patch is upstreamed.
+
+## Support This Work
+
+Our training compute is entirely self-funded. If this model is useful to you, consider supporting the lab:
+
+**[Support on Ko-fi](https://ko-fi.com/djlougen)**
+
+* * *
 
 ## Model Details
 
@@ -41,8 +49,31 @@ The fork now includes both per-layer `layer_types` support and an **SSM tensor p
 * **Context length:** 131,072 tokens
 * **License:** Apache-2.0
 
-## Support This Work
+## Usage
 
-Our training compute is entirely self-funded. If this model is useful to you, consider supporting the lab:
+### Build the patched llama.cpp
 
-**[Support on Ko-fi](https://ko-fi.com/djlougen)**
+```bash
+git clone https://github.com/DJLougen/llama.cpp.git
+cd llama.cpp
+git checkout rys-qwen35
+cmake -B build -DGGML_CUDA=ON -DCMAKE_BUILD_TYPE=Release
+cmake --build build -j
+```
+
+Drop `-DGGML_CUDA=ON` for a CPU-only build. The patch touches the GGUF loader; backend selection is independent.
+
+### Download + run
+
+```bash
+./build/bin/llama-server \
+    -m ornstein-3.6-27b-rys-q4_k_m.gguf \
+    --host 0.0.0.0 --port 8080 \
+    --n-gpu-layers 99 --ctx-size 131072 \
+    --flash-attn on --jinja \
+    -ctk q4_0 -ctv q4_0
+```
+
+## License
+
+Apache 2.0